The purpose of this study is to establish limits for stimulus parameters to avoid tissue damage and electrode damage. We will establish safe limits for toxic products and minimize the charge injected into corrosive processes. Neural prostheses can be designed to effect controlled neurotransmitter release by selectively opening and closing voltage gated ion channels on select nerves, which opens numerous opportunities to restore or replace missing or impaired organ function. Selective control of these ion channels can be realized by precise positioning of electric fields and manipulation of the nonlinear properties of gates in these channels. Precise positioning of the electric fields involves reducing the physical size and area of the electrode. Manipulating the nonlinear properties of voltage gated ion channels involves stimulus pulse durations that are five to ten times longer than those used to simply create a propagated action potential. Small electrodes and long duration stimulation pulses both increase the likelihood of electrochemical reactions on the electrode surface that create toxic products and accelerate corrosion rates. Toxic reaction products kill the cells that are to be activated, and corrosion shortens the useful lifetime of the electrode. Both combine to shorten the useful life of a neural prosthesis. This study will test the hypothesis that electrochemical reaction products created at the electrode surface during the cathodic phase of stimulation can damage cells if generated at a sufficient rate. Limits will be established for the charge that can be safely injected into toxic reactions. A method will be developed to specify an imbalanced charge stimulation waveform, which is limited in the cathodic phase by charge injection into toxic electrochemical products, thus preventing tissue damage, and in the anodic phase by limiting the charge to levels required to return the electrode to the resting state, thus preventing corrosion. The results of these experiments will improve our understanding of tissue damage, and will improve our ability to deliver a safe neural prosthesis for clinical use, with fewer animal experiments. In addition to allowing design of a safe neural prosthesis by limiting reaction products, this work may lead to productive use of reaction products such as tumor destruction and intentionally induced ischemia. This work is expected to lead to the identification of the toxic product now believed to by an oxygen free radical created during oxygen reduction.